Font Size

100% 150% 200%

Background Colour

Default Contrast
Close Reset

The Howe

Broch (Iron Age), Chambered Cairn (Neolithic), Souterrain (Prehistoric)

Site Name The Howe

Classification Broch (Iron Age), Chambered Cairn (Neolithic), Souterrain (Prehistoric)

Alternative Name(s) Howe Of Howe; Hillock Of Howe

Canmore ID 1731

Site Number HY21SE 41

NGR HY 2759 1092

Datum OSGB36 - NGR

C14 Radiocarbon Dating


Ordnance Survey licence number 100057073. All rights reserved.
Canmore Disclaimer. © Copyright and database right 2020.

Toggle Aerial | View on large map

Digital Images

First 100 images shown. See the Collections panel (below) for a link to all digital images.

Administrative Areas

  • Council Orkney Islands
  • Parish Stromness
  • Former Region Orkney Islands Area
  • Former District Orkney
  • Former County Orkney

Archaeology Notes

HY21SE 41 2759 1092.

(HY 2759 1092) The Howe (NR)

OS 6"map, Orkney, 2nd ed.,(1903).

The Hillock of Howe is a burial mound in which were found a Viking black glass linen-smoother and several other objects.

J G Marwick 1928; H Shetelig 1940.

The Howe is a conspicuous mound, almost entirely turf-covered and partly composed of burnt material. It is often refered to as a broch but there is no conclusive evidence of this, although there is distinct evidence of stone con- struction.

Partial excavation 'many years ago' produced a considerable number of relics (some of which were presented to the NMAS) and a Norse glass linen-smoother, probably Viking, in the possession of J G Marwick.

A modern cairn has been erected on the highest point of the mound as a navigational mark.

Proc Soc Antiq Scot 1889 (Donations); G Petrie 1890; J Cursiter 1923; J G Marwick 1928; RCAHMS 1946, visited 8 August 1928.

National Museum of Antiquities of Scotland (NMAS) Accession card index; NMAS attendants' catalogue.

There is no conclusive evidence to suggest that the Howe, a grass-covered mound of burnt earth and stones, is the remains of a broch. It measures c. 40.0m. in diameter and is 4.5m. high: the modern cairn is now but a small pile of stones 0.2m. high.

The linen-smoother in the possession of J G Marwick was presented to the Stromness Museum (Accession no.119) but has since been lost.

Resurveyed at 1/2500.

Visited by OS (RD) 14 September 1964.

Excavation commenced in 1978, and still continuing until complete removal, has revealed that the site consists of an Iron Age ring-fort overlain by a broch and a later Pictish settlement on the S side of the broch. Investigation of the rampart of the fort has revealed a 1m thick faced rubble wall capped with clay, in its final form 3.5m wide and 2m high. The broch survived to a height of c.5m with wall 5.5m wide and containing an intra-mural staircase. The entrance was in the SE. This broch structure overlay a thin-walled primary broch and contemporary with which was a souterrain sealed by later floors. The broch had become a workshop with 2 kilns, charcoal oven & associated hearths. S of the broch workshop structures were revealed directly over primary outhouses. Below the broch tower is evidence of a much earlier structure.

The initial Pictish occupation made use of broch-age structures to the S, but then expanded to the W; a variety of house types was evident. A curious rectangular structure on top of the mound may well prove to belong to the Norse period.

Finds range from the Bronze Age to Norse and Medieval.

J Hedges and B Bell 1979; S Carter et al 1980.

In 1981 further excavation confirmed the presence of a possible chambered tomb, associated with a hornwork bank with an outer ditch, or overlying it. There appear to be 3 side cells off the main chamber which was entered by a passage 7m long with steps down to the main chamber.

At a later stage the centre of the tomb mound was cut away to ground level, and levelled with clay over the chambers which were re-roofed and used as a souterrain for the round house (with radial partitions) which was built over the tomb. The tomb passage was used as a drain.

Directly over all this was built the 2-phased broch, excavated in 1979-80.

Finds included a fibula, gilded bronze needle, pottery, and skeletal remains.

B Bell and D Haigh 1981.


Excavation (1978)

Excavation in 1978.

Source: J Hedges and B Bell 1979; S Carter et al 1980.

Excavation (1981)

In 1981 further excavation confirmed the presence of a possible chambered tomb, associated with a hornwork bank with an outer ditch, or overlying it. There appear to be 3 side cells off the main chamber which was entered by a passage 7m long with steps down to the main chamber.

At a later stage the centre of the tomb mound was cut away to ground level, and levelled with clay over the chambers which were re-roofed and used as a souterrain for the round house (with radial partitions) which was built over the tomb. The tomb passage was used as a drain.

Directly over all this was built the 2-phased broch, excavated in 1979-80.

Finds included a fibula, gilded bronze needle, pottery, and skeletal remains.

B Bell and D Haigh 1981.

Orkney Smr Note (September 1987)

Bone handle of an implement, two whorls and a disc of

sandstone, portion of a vessel of steatite and part of a tubular

handle of an earthenware Porringer - from Howe. [R1]

The mound is often referred to as a broch but there is no

conclusive evidence of this. [R2]

A Norse glass linen-smoother, probably Viking, in the

possession of J G Marwick. [R3]

The Hillock of Howe is a burial mound in which were found a

Viking black glass linen-smoother and several other objects. [R4]

The Howe is a conspicuous mound, almost entirely turf-covered

and partly composed of burnt material. Partial excavation 'many

years ago' produced a considerable number of relics. A modern

cairn has been erected on the highest point of the mound as a

navigation mark. [R5]

There is no conclusive evidence to suggest that the Howe is

the remains of a broch. It measures about 40.0m in diameter and

is 4.5m high. OS visit Sept. 1964.

Completely excavated to ground level by North of Scotland

Archaeological Services; publication in progress by B Smith.

Remains of a Maeshowe-type chambered tomb were overlain by

(unrelated) broch with surrounding settlement; the broch was

preceded by a proto-broch structure, and occupation extended into

the early Pictish Period.

Information from Orkney SMR (RGL) Sept. 1987

Publication Account (2002)

HY21 6 THE HOWE (‘Hillock of Howe’, ‘Howe of Howe’, ‘Cairston’)


Roundhouse, dun, solid-based probable 'broch' and later Pictish village on farmland in Stromness. This was originally a conspicuous green mound which was classed as a probable broch by the Commission [4] having been partially excavated in the 19th century when it produced many finds some of which were presented to the National Museum [3, 8]. No coherent evidence for stone buildings seems to have been recovered at that time. Systematic excavations (under the leadership of John Hedges to begin with) took place between 1978 and 1982 [7] and revealed a complex sequence of dry stone structures -- dating from Neolithic to Pictish times -- and including two successive broch-like buildings (excavation visited in July 1980).

In terms of scale and thoroughness this was one of the largest and most expensive excavations ever undertaken in Scotland, and it has provided a sequence of structures and artefacts which is only matches in Iron Age Scotland by that at Jarlshof (HU30 1). The information gained seems likely to revolutionise our understanding of the north Atlantic Iron Age [7]. A point of major importance is that, since the site was to be destroyed, a complete excavation could be carried out; there was no problem about removing the later structures to examine the earlier ones fully.

The work has been reported on in great detail [13] and the author has reviewed this work extensively [14]. A summary is therefore seems sufficient here, together with an indication of where the author' understanding of the site differs tentatively from that presented in the report.

1. The excavations

1.1 Continuity with the past?

The Iron Age buildings were planted on the ruins of two superimposed Neolithic chambered tomb most of the details of which may be ignored here. However three architectural features of the Neolithic structures seem to have influenced the design of the later buildings, to such an extent that it may be argued that there was, if not complete cultural continuity between the two periods (in idea rendered dubious by the completely different material culture of the later periods), then at least a major attempt by the Iron Age settlers to refer back to the much older religious site in several important ways. It may be further argued that the Iron Age site itself may have been -- both because of its siting and because of the deliberate links with the Neolithic architecture -- more ceremonial than defensive [14]. There are more comments on this theme in the Discussion section.

The first link with the design of the Neolithic site is basic. The stone passage grave (the second tomb) was contained within a mound of clay-like soil which was itself surrounded by a ditch cut into that same subsoil. It is possible that the passage grave survived intact until the first Iron Age occupation (Phase 4), or it may have been in ruins long before; however the builders of the Phase 5 roundhouse (below) sited the structure on the remains of the clay mound (after the passage grave had been largely removed) and cleared out the ditch. The unstable clay foundation resulted in collapses of the roundhouse wall but this did not deter subsequent builders from using exactly the same unsuitable site for their much heavier buildings.

In the second place the long, almost straight entrance of the passage grave seems to have determined the positioning and alignment of the entrances of the three successive round Iron Age buildings which used the same passage paving for their entrances. Thus the circular buildings themselves were laid out around very similar centres.

The third link with the past is the chamber below ground level which formed part of the passage grave, though how it was roofed then is not known. This remained open as a sort of cellar, and quite free of occupation debris or other clutter, inside the round buildings throughout most of the Iron Age occupation of the site except Phase 8. At one point it was extended and turned into something more like an Iron Age souterrain, with upright stone slabs supporting the roof slabs. The phenomenon of elaborate underground chambers in Iron Age Orkney -- mostly 'wells' inside or next to brochs but including the extraordinary Minehowe site on a low hilltop -- seems to be unique and may require a ceremonial rather than a practical explanation.

1.2 Phase 5

After some ill-understood early Iron Age activity the first reasonably well defined building detected was a defended stone roundhouse with a massive stone wall which had been almost completely destroyed by later activity; this was apparently protected by, and therefore contemporary with, an outer defence work consisting of a clay-cored rampart with a rock-cut ditch outside it (which, as noted, are the remains of the clay mound and surrounding ditch of the Neolithic passage grave). These outworks are referred to as the “ring-fort” in some of the interim reports. The roundhouse -- with an internal diameter of c. 7 m -- had an underground cellar, or earth house, attached which was reached by a vertical shaft in it floor. Two radiocarbon dates in the 5th century b.c. were obtained for this occupation; they were 430 +/- 50 b.c. (GU 1799) and 455 +/- 70 b.c. (GU 1789).

1.2 Phase 6

Apparently the roundhouse of Phase 5 eventually collapsed, the outer face sliding down the slope of the mound and the inner face shifting as the clay below settled. It was dismantled and replaced by another “roundhouse or broch”; this building had a wall 3.5 m thick which survived to a height of up to 2 m. It contained two intra-mural stairs, each a short clockwise flight. The entrance was 2.1 m long and had some kind of “open-backed” cell on either side of it. The interior was divided by stone partitions into (1) an outer zone of paved radial rooms or bays and (2) a central area with the hearth. The earth house of Phase 5 remained in use. Post-holes in a rough circle were discovered belonging to phase which could have supported a roof, with perhaps also a timber gallery. A C-14 date of 305 +/- 95 b.c. (GU 1758) was obtained for this phase [7].

1.3 Phase 7

The interior of the roundhouse was cleared and the area between it and the outer rampart was levelled. A 'broch' with a wall 5.5 m thick was built over the remains of the roundhouse, the old entrance being re-used; the cell and intra-mural stair in the west were rebuilt. However early in its construction the new building collapsed in the west -- presumably because of the unstable clay foundations -- and this caused the cell and the staircase there to lose their roofs, though much of the inner face of the broch remained intact. The outer face was rebuilt and the cell was re-floored and buttressed.

The new central court consisted of a central area 4 m in diameter containing traces of a hearth in the middle; the outer area comprised, in the north and E, the curved passage housing the earth house shaft and, on the south and W, three bays -- two “lidded” (presumably roofed with stone slabs) and one open. The west cell and mural stair were blocked up because of instability and were replaced by an internal stair built at the end of the curving passage. Six stone houses were built next to the broch on the E, south and west sides (on either side of the broch entrance) and within the outer rampart; they were apparently contemporary with the first phase of the broch's occupation rather than being secondary outbuildings. There was also a paved courtyard in front of the broch entrance and an external celled doorway was added to the entrance. Several C-14 dates were obtained for material of Phase 7: 120 +/- 50 b.c. (GU 1750) for the end of the main broch village, AD 15 +/- 55 (GU 1788) for an early episode of burning inside the broch, 25 +/- 55 b.c. (GU 1786) for a late burning inside the broch and AD 280 +/- 65 (GU 1787) for a workshop floor in the NE outbuilding.

1.4: later Phase 7 and Phase 8:

The outbuildings were modified in various ways as time went on and the outer face of the broch tower collapsed again severely in the west and various shoring up operations took place. The interior was relatively undamaged and was re-occupied as a workshop equipped with kilns or ovens and also used for pottery and stone tool manufacture.

The final collapse of the broch wall blocked its entrance and access thereafter was over the wall head and down the internal staircase of Phase 7. Some further workshop activity took place but more rubble falls caused the ruined building to be abandoned. Further activity took place in the outbuildings and continued on into Pictish times when new dwellings were built on top of the original outbuildings. Two C-14 dates were obtained for this latest period -- AD 385 +/- 45 (GU 1749) and AD 500 +/- 50 (GU 1757).

1.5 Finds: (1) pottery

Only a brief summary of the succession of material cultures can be given here [13]. The appendices in the report -- which should be studied carefully by those interested -- provide numerous parallels for the pottery and the other artefacts. Striking is the number of exotic-looking bronze pins and brooches.

Late Bronze Age / early Iron Age pottery. It is difficult to get a clear idea of the wares present at Howe (except from the drawings) because of the habit of describing the sherds by their separate characteristics, such as form, fabric and decoration. Relatively few finds were recovered from Phases 4 and 5 and in particular no clearly carinated pottery, like that found in the earliest roundhouses in Orkney (HY20 4, HY44 8 and HY41 5), was identified. One such sherd is claimed: [13, Illus. 139, no. 6956] but the 'carination' is barely perceptible.

However there is another characteristic LBA/EIA pottery from mainland Scotland, the distribution of which extends into Caithness, of which a few sherds turned up in Phase 5 deposits at Howe [13, Illus. 141, nos. 6410 and 4400). This is a smaller and more delicate vessel, or jar, of the plain, gritty Dunagoil ware most of which are in the form of barrel-shaped urns. Other parallels are listed [13, 241]. There is also something like a roughly made and crude form of a small Everted Rim jar [13, Illus. 141, no. 7083].

Middle Iron Age pottery. Most of the sherds from the later strata came from Phase 7, the occupation of 'Broch 2' [13, Table 70]. Very rare are the standard Everted Rim jars often found on Shetland and Hebridean sites; the top two pots in Illus. 152 [13] might be examples. Many of the diagnostic sherds appear to come from vessels with incised decoration [13, Illus. 146 and 148-49) quite similar to Vaul ware vases from the Western Isles. These are rare in Orkney; sites like Gurness and Midhowe yielded mainly plain ware and until now the only two sites which have given incised ware in any quantity have been Ayre (HY40 1) and Lingrow (HY40 2).

Another quite common style is a version of the Everted Rim jar with neck-band decoration, common in Shetland, although at Howe the sharply everted rim is hardly present [13, Illus. 144]. Again a variety of possible parallels is offered [13, 250].

14. Finds: (2) other artefacts

As with the pottery finds of other types are rare before Phase 7 [13, Table 57].

Metal: a number of unusual iron objects were found [13, Illus. 130], including a small pair of tweezers, 2 decorated pins, an almost intact razor blade set in a polished bone handle [ibid. no. 5319], some knife blades (including one with an almost complete tang with a small bone pommel [ibid. Illus. 130, no. 5165]) and two small chisels, one of which may have been a chaser for cutting ornamental patterns [ibid., Illus. 130, no. 4822 and Illus. 131]. Similar implements were found at Leckie broch, Stirlingshire (NS69 3). There is also a late disc-headed pin comparable to the one found at Burray East.

Potentially important too are the two examples of long iron pins with ball-shaped heads -- usually of bone -- of rare type. In the first case only the head remains, the iron shank having become detached or rotted away; the head is of polished jet [13, Illus. 107, no. 294]. The other (from Phase 7) is complete and the head is an oval pale glass ball [13, Illus. 90, e].

A bronze ring-headed pin of north British type (with a projecting head) comes from deposits of Phase 5/6 and is thought indirectly to date the building 'Broch 1' to perhaps the 1st or 2nd centuries BC, held to be the early Iron Age; however the type has a very long life span [14, 24-5] and others came from later deposits [13, Illus. 133 and 134], so relying on such pins for a date is unwise.

The later such pins include one which strongly resembles the 'elbow studded' La Tene form (without the projecting head) which can be as early as the 3rd/4th centuries BC in the south.

Other bronze pins (Phase 7) include a long one with a zoomorphic head the shaft of which has been twisted longitudinally [13, Illus. 133, no. 4314]; a good parallel (broken and without a head) comes from Dun Ardtreck, Skye (MacKie 2002, Illus. 25, no. 44]. The 4th/5th century AD date for the type (Fowler 1963, 103) supports a late date for the middle Iron Age occupation at the Skye site.

There two complete penannular fibulae are of the zoomorphic type which are thought to have developed in the north in the 4th and 5th centuries AD, perhaps because of the influence of zoomorphic pins like the one just mentioned. The terminals are almost identical to that of the pin and it is extremely rare to find both types of ornament together.

Two highly unusual brooches are the one formed like an insect with tinned wings spread out [13, Illus. 135] and a La Tene 1 fibula of the native British 'Swallowcliffe' sub-type found in late Phase 8 deposits. The latter must be an heirloom as the type, of southern manufacture, can hardly be later than the 3rd century BC.

The insect brooch seems to be of Roman make and is likely to have been made in the 2nd century AD.

What is probably a fragment of a bronze spiral finger ring was also found [13, Illus. 132, no. 205]. In spite of the accompanying comment (which relies on Clarke 1971) the writer knows of no such rings in the north which are earlier than the middle Iron Age (that is, earlier than about the 2nd/1st centuries BC). Whether the ones found at Gurness are of Norse date (as is also suggested, following Hedges et al. 1987, 87) is highly debatable, although of course it is possible that Iron Age finger rings were in use as heirlooms for a long time.

A pair of small bronze tweezers was found together with two bronze needles [13, Illus. 136, nos. 168 and 4414]. The tweezers are similar to the iron ones (above) and, being rare in the north, could be copies of Roman ones. One of the needles resembles that found at Dun Ardtreck, Skye, which is thought to be Roman on the basis of the groove below the eye (MacKie 2002, Ills. 28 and 25, no. 41]. The Howe example by contrast does not show any groove in the drawing.

Glass objects included pieces of Roman vessels [13, 234 and Illus. 138], and a Roman intaglio from a late (Phase 9) horizon; the latter is carved with an eagle and was probably made in the middle of the 2nd century. Like many other Roman objects it doubtless became an heirloom. Native glass included several beads, two being large fragments of thick ring beads with coloured inlay [13, 234-5 and Illus. 138]; a very similar example was found at Dun Mor Vaul [NM04 4]. There were also two small, opaque, yellow annular beads of the standard middle Iron Age type, found in many brochs, and an unperforated dumb bell object, presumably a bead. The last is very late, from Phase 9, but a similar object was found at Leckie broch in Stirlingshire where it was well dated to the late 1st or the 2nd century AD (MacKie 1985).

Of the stone implements the most interesting are perhaps the rotary querns of which the new drawings made by the writer are given again here (MacKie 1998, 26-31). Of particular interest are the two with radial handle slots -- a southern English and north-west French type. Saddle querns were also found which again suggests that both kinds of grinding mill were in use simultaneously at some site in the middle Iron Age; indeed the rotaries were out-numbered by the saddle stones in late Phase 8 [13, 207]. There is a possibility that one fragment of a rotary quern dates back to the early Iron Age but it is not illustrated [13, Table 53].

There are also four fragments of bracelets, two of jet and two of stone, of a type commonly found on Scottish Iron Age sites [13, Illus. 107];

Bone artefacts were very varied but the standard types include both long-handled 'weaving' combs and late Iron Age composite combs [13, Illus. 101 and 100]. There were also an unusual, single piece, single-sided rectangular hair comb [13, Illus 100, no. 4907 and Illus,. 90, a].

Two centrally perforated 'netting needles' with a point at each end are more likely to be fish gorges (not 'gouges') [13, Illus. 95, nos. 3580 and 2139]. Any kind of needle by definition is long, fairly slender and has the hole for the thread near one end.

2. Discussion

Only a few aspects of this immensely complex and important site can be commented on here; the writer has already offered an overview of its significance elsewhere [14].

The question of the nature of 'Broch 1' and 'Broch 2' has been discussed at some length there; neither building -- massive, round and broch-like in proportions though they are -- produced any evidence for the hollow-walled architecture of the tower brochs. Neither was there clear dating evidence that 'Broch 1' belonged to the early Iron Age and thus provided evidence to support the evolution of the brochs in Orkney from the roundhouses of that period. 'Broch 1' could not really be pushed back before the 1st century BC or the 1st AD.

Indeed the unstable clay underlying the central buildings resulted in the partial collapse of all three round buildings erected on it; moreover the core of the wall of the massive 'Broch 2' was also partly of clay rendering it quite unsuitable to be turned into a high tower. There can really be little doubt that Howe broch 2 was an 'imitation broch' -- a relatively low-walled structure the designers of which copied some aspects of broch architecture but could not build -- or did not wish to build -- a high tower.

The second point has also been dealt with at some length [14, 33 ff.]. It is possible to make out a strong case that Howe throughout the Iron Age was as much a ceremonial as a defensive site, and that its situation on top of a wrecked Neolithic passage grave (resulting in very unstable foundations) shows that it was more important to have physical continuity with an important monument of what was even then a very remote period than to construct an effective stronghold.

A study of the finds does seem to confirm this conclusion. Although the pottery on the whole seems to have indigenous Orkney origins there are enough rare and unusual metal ornaments to suggest that the site had regular long distant contacts with much more southerly regions of the British Isles.

Taking all these points together it may not be too fanciful to suggest that Howe was a very ancient holy site with high prestige and that high status pilgrims -- perhaps priests or chiefs or both -- came to it regularly from far away for important religious and political reasons.

Sources: 1. OS card HY 21 SE 41: 2. P.S.A.S 23 (1888-89), 238 (finds): 3. Marwick 1928: 4. RCAHMS 1946, 2, 324, no. 921: 5. J Hedges et al. in Discovery and Excavation 1979, 24: 6. Hedges and Bell 1980: 7. S.Carter et al. in Discovery and Excavation 1980, 24: 8. Bell, B and Haigh, D in Discovery and Excavation 1981, 25: 9. Haigh and Smith 1982: 10. S. Carter et al. 1984: 11. Curister 1923, 50: 12. Hedges et al. 1987, 92-3: 13. Ballin Smith (ed.) 1994: 14. MacKie 1998.

E W MacKie 2002


MyCanmore Image Contributions

Contribute an Image

MyCanmore Text Contributions